Syslinux - Jan 2009 - "UI" keyword, pc_default and getline

There is a problem with "pc_default" if the following is present in
the config file (line order does matter):

UI gfxboot.com init
DEFAULT memtest

"pc_default" is called on the "DEFAULT" keyword, but
"getline" is skipped, so the code now tries to
parse the keywords parameter. The config line should be skipped in this case.

This could be fixed by converting "skipline" to a function and calling
it from "pc_default".

- Sebastian

diff --git a/core/parsecmd.inc b/core/parsecmd.inc
index edca0b7..27e115f 100644
--- a/core/parsecmd.inc
+++ b/core/parsecmd.inc
@@ -97,12 +97,18 @@ getcommand:
 .eof:          stc
                ret

-.skipline:     cmp al,10               ; Search for LF
+.skipline:     call skipline
                je .find
-               call getc
                jc .eof
                jmp short .skipline

+skipline:      cmp al,10               ; Search for LF
+               je .end
+               call getc
+               jc .end
+               jmp short skipline
+.end:          ret
+
                section .data
 err_badcfg      db 'Unknown keyword in configuration file: ',0
 err_noparm      db 'Missing parameter in configuration file. Keyword:
',0
diff --git a/core/parseconfig.inc b/core/parseconfig.inc
index fcf18d9..57ee5a5 100644
--- a/core/parseconfig.inc
+++ b/core/parseconfig.inc
@@ -26,7 +26,9 @@ pc_default:   cmp ax,[DefaultLevel]
                mov di,default_cmd
                call getline
                mov byte [di-1],0               ; null-terminate
-.skip:         ret
+               ret
+.skip:         call skipline
+               ret

 ;
 ; "ontimeout" command

Sebastian Herbszt wrote:
> There is a problem with "pc_default" if the following is present
in the config file (line order does matter):
> 
> UI gfxboot.com init
> DEFAULT memtest
> 
> "pc_default" is called on the "DEFAULT" keyword, but
"getline" is skipped, so the code now tries to
> parse the keywords parameter. The config line should be skipped in this
case.
> 
> This could be fixed by converting "skipline" to a function and
calling it from "pc_default".
> 
Thanks!

This patch didn't apply for me so I rewrote it, but the idea is the same.

	-hpa

-- 
H. Peter Anvin, Intel Open Source Technology Center
I work for Intel.  I don't speak on their behalf.